Prompt Engineering State-of-the-Art: Prompting as Behavioral Engineering for Google Gemini
Year: 2026
Type: article
Abstract: This study aims to establish a unified, industry-standard taxonomy of prompt engineering techniques and evaluate their effect as behavioral triggers for LLMs. Despite proliferation in research, the prompt engineering field remains fragmented, lacking a consolidated vision that bridges academic theory with industrial application. This research synthesized 228 academic methods and 47 industry-validated protocols to identify a state-of-the-art list of 20 recurrent prompt engineering methods. Using Google Gemini, these techniques were tested, measuring variations in computational behavior, specifically latency and token consumption. Results demonstrate that prompting functions as a form of behavioral engineering; while all 20 techniques yielded correct answers, they triggered statistically significant variances in computational resource demand. Nonparametric analysis confirmed that each technique elicits a distinct behavioral profile, where latency and token consumption are indicators of the model’s internal cognitive process.The study concludes that prompting acts as behavioral stimuli, providing a unified framework for high-efficiency human–computer interactions.
Institution: Tecnológico de Monterrey
Language: English
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7204169676